Adirondack Lakes and Ponds
2023-04-10
Here I provide an overview of data available to analyze for the SCALE project.
Future SCALE sampling efforts will be tiered with different intensity levels depending its level of priority. All lakes will be sampled using remote sensing, 300 will be sampled once for basic characteristics, 50 will be sampled at intermediate levels (more variables, several sampling trips), and 25 will be sampled intensely (i.e., monitoring buoys, multiple sampling trips, etc) Our goal is to categorize which lakes should be sampled with which degree of intensity. We want to ensure that each sampling intensity level covers a range of waterbody types and characteristics found in the Adirondacks.
1 Waterbodies in the Adirondack Park
The first step is to determine the basic characterisitics of all waterbodies in the park. To do that we turn to the National Hydrography Dataset (NHD). We look at all waterbodies in the NY region, and then filter out those that lie within the boundary of the Adirondack Park (Figure 1.1).
Figure 1.1: Waterbodies in NY (grey) with those within the ADK park highlighted (blue)
The NHD lists nearly 11215 waterbodies in their dataset. Of those, only 1750 have GNIS ID numbers and names. A fun fact is that the most common name among lakes and ponds in the Adirondack Park is Mud Pond.
These lakes and ponds range from as small as 24 m2 to as large as 1121 km2. The NHD also includes elevation for 1115 of the lakes and ponds. The lowest elevation is an unnamed lake (0.12 km2) at 28.3 m, while the highest elevation in the dataset is for Woodworth Lake (0.138 km2) at 1680 m.
The quantiles for lakes and pond surface area (ha) are: 0, 0.07, 0.11, 0.2, 0.29, 0.41, 0.7, 1.2, 2.55, 7.4, 1.1214555^{5}
90% of lakes and ponds are less than 7.4 ha. Of those that are larger the quantiles (0.91-0.99 by 0.01) are: 8.5, 9.9, 11.9, 14.4, 18.5, 24.4, 36, 56.6, 129.2
From the NHD we can also use data on watershed boundaries. The NHD includes multiple watershed boundaries based on several hydrological units. We are looking at HU10 (Figure 1.2), which includes a major waterbody.
Figure 1.2: Waterbodies in ADK with watershed boundaries at the HU10 level
1.1 Land Cover
To determine the types of land cover around each waterbody we turn to the National Land Cover Database (NLCD; Figure 1.3). The NLCD gives land cover estimates for 30 m by 30 m grid cells.
Figure 1.3: Land cover in the Adirondack Park
Over the whole park, the most common land cover type is deciduous forest (46%), followed by evergreen forest (21%), mixed forest (12%), and woody wetlands (9%; Figure 1.4). About 5.6% of the park is open water.
Figure 1.4: Land cover percentages in the Adirondack Park
2 Remote Sensing
NLDAS data here
3 Historical Monitoring
3.1 Brief review of historical monitoring
Adirondack water bodies have been the subject of numerous long- and short-term monitoring programs.
3.2 Ice cover
For 8 lakes in the Adirondacks we have a record of ice-in and ice-out from 1874 through 2022. Ice-out records show a generally declining trend, with ice-out occurring earlier each year (Figure 3.1). There does not, however, seem to be any apparent trend in the duration of ice cover in these lakes, though the variance in duration may be increasing (Figure 3.2).
Figure 3.1: The ice out date for each year for each of 8 monitored lakes
Figure 3.2: The period of ice cover for each of 8 monitored lakes. Zero indicates the start of the year (Jan 1), with negative values being days prior to the start of the year.
3.3 Adirondack Lake Survey
The Adirondack Lake Survey (ALS).
3.3.1 1984-1987
The 1984-1987 ALS data can be found on Figshare here. It includes tables on location, status, morphometrics, substrate, watershed characteristics, inlets and outlets, macroinvertebrates, aquatic vegetation, physical parameters, gear type for fishing, individual and community fish data, water chemistry, liming, and reclamation.
Figure 3.3: ALS lake elevation, volume, watershed area, surface area, max depth, and mean depth
Figure 3.4: ALS temperature profiles by month
Figure 3.5: August thermocline depth by surface area
Figure 3.6: ALS lake dissolved oxygen profiles by month
Figure 3.7: ALS lake chemistry at 1.5 m
Figure 3.8: Chloride in the upper 1.5 m of ADK lakes and ponds
Figure 3.9: Biplot of substrate type PCA on 1467 lakes/ponds
Figure 3.10: Biplot of PCA on relative abundance of fish species in 1023 lakes
3.3.2 1992 ALTM
Figure 3.11: DOC over time in ALTM lakes
Figure 3.12: Chloride over time in ALTM lakes
Figure 3.13: Chlorophyll over time in ALTM lakes
3.3.3 1994-2005 Fish
The 1994-2005 ALS fisheries data can be found on Figshare here.
3.3.4 2008-2012 Fish
The 2008-2012 ALS fisheries data can be found on Figshare here.
3.4 AEAP
There are 28 lakes in the AEAP dataset, obtained through the R package adklakedata. The package includes datasets on chemistry, crustacean zooplankton, rotifer zooplankton, phytoplankton, lake nutrients, secchi depth, temperature and dissolved oxygen profiles, metadata, and lake-specific meteorology.
Figure 3.14: DOC over time in AEAP lakes
3.5 Citizen’s Statewide Lake Assessment Program
The Citizen’s Statewide Lake Assessment Program (CSLAP) is a volunteer-based monitoring program for lakes around New York State. CSLAP is managed by the NY Department of Environmental Conservation (NYDEC) and the New York State Federation of Lake Associations (NYSFOLA). Citizen volunteers collect samples that are analyzed by state (or state-approved) labs.
Of the 259 lakes in the program there are about 52 lakes in CSLAP within the Adirondack Park, with 72 total sampling locations (Figure 3.15). These data (available on EDI here) run from June of 1986 through July 2012, though additional data may be available elsewhere.
Figure 3.15: New York lakes involved in the CSLAP program, with the ADK park highlighted.
These CSLAP data include information on each lake and sampling point, as an ID number assigned by NYDEC, an official lake name, and a NYS pond number. Other metadata include a NYS volume code, surface area, bottom depth, sample date, sample depth, sample type, latitude, and longitude. Samples are typically taken either 1.5 m below the surface or 1.5 m above the bottom, though other sample depths are included. The chemistry data include total phosphorus, NOx (nitrite and nitrate), NH4, total nitrogen, and chlorophyll. Samples also include secchi depth, temperature, and color.
The surface area of lakes in CSLAP within ADK park range from 0.1 km2 to 114 km2
Figure 3.16: Distribution of lake surface area in CSLAP lakes within ADK park
A look at general trends across all ADK CSLAP lakes suggest that nutrients (TN, TP) are remaining relatively flat, temperature is rising, while chlorophyll and color are decreasing and sechhi depth is getting deeper (Figure 3.17). The trend in secchi and color seem to buck trends observed in more generalized monitoring. which typically indicate shallower secchi depth due to darkening lakes (with higher DOC as a result of reduced acidification).
Figure 3.17: Secchi depth, color, total phosphorus, total nitrogen, and chlorophyll a in CSLAP lakes within the Adirondack Park (at 1.5 m for all but Secchi). Lines are output from a Generalized Additive Model.
3.6 Lake multi-scaled Geospatial and temporal database
The LAke multi-scaled GeOSpatial and temporal database (LAGOS-NE-LIMNO v1.087.3) is a data package that includes information on lakes and reservoirs in 17 lake-rich states in the Northeastern and upper Midwestern United States. The LIMNO part of the LAGOS-NE database includes in-situ measurements of lake water quality over the past 30 years, taken from federal, state, tribal, and non-profit agencies. Overall it covers 2600-12000 lakes (dependent on variable).
Figure 3.18: Location of lakes in the LAGOS-NE-LIMNO dataset
The max depth in LAGOS-NE-LIMNO lakes ranges from 0.6 m to 50 m, and mean depth ranges from 0.3 m to 21.3 m.
Figure 3.19: Mean and max depth in lagos lakes
We can also look at trends in multiple values from the LAGOS-NE-LIMNO dataset.
Figure 3.20: Secchi depth, color, DOC, chlorophyll a, total phosphorus, and total nitrogen in LAGOS-NE lakes within the Adirondack Park. Lines are output from a Generalized Additive Model.
3.7 Combining
I started with the NHDplus and LAGOS-NE data. Then I identified AEAP lakes and ALSC lakes by their legacy id value in LAGOS. For EPA TIME and NY CSLAP I linked the nhdid value in LAGOS with the Permanent_ id in NHDplus. Not all ponds/lakes identified in the ALSC database were found in the LAGOS database, so there may be some missing in the full dataset here. The ALSC location dataset included 4833 unique pond numbers, but only 1221 unique ponds/lakes are identified in LAGOS from that survey (from metadata seems like I missed 45?). We also inlcuded identifiers for lakes in the current SCALE pilot study.
We have included the NHDid (Permanent_), NHDPlusID, GNIS_ID, GNIS_Name, PONDNO, PONDNAME, and Lagos Name, though many of the names are redundant. We estimated the elevation of each lake in the NHD dataset using a digital elevation model, averaging cells that overlap the lake shape.
The HydroLAKES dataset includes estimates of depth, volume, shoreline characteristics, residence time, discharge and watershed area for most lakes >10 ha. Of the 889 lakes >= 10 ha, depth estimates are inlcuded for all but 155.
Figure 3.21: Mean chlorophyll across all measurements for each lake
Figure 3.22: Mean phosphorus across all measurements for each lake
Figure 3.23: Mean nitrogen across all measurements for each lake
Figure 3.24: Mean secchi, carbon, and color across all measurements for each lake
Figure 3.25: Trend in water clarity across 790 lakes in ADK
4 Modelling
4.1 Generic lake models
We simulate lakes with systematically varying characteristics to determine how max depth, mean depth, surface area, and Kd interact to determine thermal regime.
4.2 AEAP
We simulate the time series of 1D temperature profiles from 1992 through 2012 for each of the 28 long-term monitoring lakes to examine patterns of stratification.